A Short-Term Load Forecasting Model of LSTM Neural Network considering Demand Response

As one of the key technologies for accelerating construction ubiquitous Internet Things, demand response (DR) not only guides users to participate in power market operations but also increases randomness grid and difficulty load forecasting. In order solve problem rough feature engineering processing low prediction accuracy, a short-term forecasting model LSTM neural network considering is proposed. First all, view strong complexity input features, weighted method used process multiple features strengthen contribution effective tap potential value features. Secondly, an improved genetic algorithm (IGA) obtain best parameters; finally, special gate structure selectively control influence variables on parameters perform The experimental results show that research has high accuracy application provides new way development